Soil compaction and disturbance following a thinning of second-growth Douglas-fir with a cut-to-length and a skyline system in the Oregon Cascades

Soil bulk density and disturbance were measured before and after a commercialthinning of 30-50 year-old Douglas-fir using a cut-to-length (CTL) and a skyline systemon the Willamette National Forest in the Oregon Cascades. A dual-probe nucleardensimeter was used to measure bulk density at four and eight inches. Slash depth afterthinning and disturbance category were determined at bulk density measurement points inall units. Data were analyzed using paired t-tests, Tukey-Kramer multiple comparisontests, and multiple linear regression.
Data were collected in two CTL units pre-thinning but one unit post-harvester andpost-thinning (after both harvester and forwarder). During the sampling period thinningwas completed only in the northeast corner of one unit (approximately 25 acres) beforeoperations were ended due to concerns about soil moisture conditions and damage tohauling roads. Thus post-harvester and post-thinning data and conclusions apply only tothat portion of the CTL units. In this harvested portion, slopes ranged from 0-10 percentand equipment traffic levels were estimated to range from three passes at the end of skidtrails, to 30 passes on secondary trails and well over 40 passes on primary trails. Thesevalues are considered characteristic of most of Unit 82 except for areas near streams withsteeper slopes.
Plot centers were located using an incomplete, offset grid system. A sub-sampleof bulk densities in old skid trails was also taken. The effects of harvester-only and thecombined effect of harvester-forwarder traffic on the bulk density of old skid trails andpreviously undisturbed soils was assessed. In the three skyline units, bulk density was
4
measured in the center, edge and halfway between skyline roads both before and afterthinning. In addition, detailed maps of soil disturbance were created from transectslocated both directly on top of and halfway between a sub-sample of skyline roads withineach unit.
From the randomly located plots in the CTL units, it was determined that old skidtrails covered approximately 27 and 13% of the two units, respectively. In the portion ofthe one unit where thinning was completed, it is estimated that an additional 27% of thearea was disturbed, for a total of 40% disturbance post-harvester. Further, the harvesterwas estimated to have left exposed soil on 4% of the area and mixed the mineral andorganic horizons on 9%. Post-thinning (both harvester and forwarder), new disturbancewas estimated at 25%, for a total combined disturbance of 38%. Exposed or mixed soilwas not observed at this time. The discrepancy between post-harvester and post-thinningdisturbance estimates is thought to be due to different surveyors, slightly different datasets, redistribution of slash during forwarding or sampling error.
In the skyline units, the maximum unit-wide disturbance was 1.8%, thoughgreater than 10% of some units were in skyline roads. Mixing accounted for most of thedisturbance, with exposure and rutting never reaching 0.5% on a unit-wide basis. Theareas of exposure and rutting were small, discontinuous and usually occurred within 150feet of the landing.
In comparison to undisturbed soil, bulk density was still significantly elevated onthe fifty-year-old skid trails sampled in the CTL units. The old skid trail center was 10-16% more dense than undisturbed soil at the four inch depth in both units, but only ruts inUnit 81 were also greater in density (+8%) at the same depth. At eight inches, old skid
trail ruts in both units were consistently more dense (13-15%), as well as the center oftrails in Unit 82 (+15%). In neither unit was the bulk density of old skid trail bermsstatistically different from undisturbed soil. Overall, it is estimated that 4 and 10% ofUnit 81 (4 and 8 inch depth) and 6 and 4% of Unit 82 was already in a compacted stateprior to this thinning entry. It is important to note, however, that elevated bulk densitieson old skid trails may also be explained by measurements being made on naturally moredense subsoils exposed after scalping of the surface soils during the initial harvest entry.
The harvester alone was not found to significantly compact either undisturbed soilor old skid trails (p>0.10). Harvester-forwarder traffic significantly increased the bulkdensity of previously undisturbed soil by an average of 12 (4 inch depth) and 11% (8 inchdepth, p=0.04 and 0.05), but did not change the bulk density of old skid trails.Compaction levels as a result of this entry were comparable with those of old skid trails.Overall, it is estimated that this new entry contributed an additional 25% (4 and 8 inches)to the compacted area, for an estimated total compacted area of 29-31%. The estimate oftotal compaction due to this thinning is based on the estimated area in new skid trails andan average of all observed compaction values across the width (both ruts and center) ofthe skid trail.
The bulk densities of the center, edge and zone between skyline roads afterthinning were compared to pre-thinning values, but revealed no evidence that asignificant difference existing between the four categories (90% level). There was someevidence that mixed soil with ruts greater than six inches deep were approximately 40 and45% (4 and 8 inch depth) more dense than undisturbed soil (p=0.0006 and 0.0002). Also,points along the skyline roads where over 16000 ft3 were yarded were found to be
significantly associated with a reduction in bulk density of approximately 16 and 18% atfour and eight inches in depth (p=0.022 and 0.008).